The recovery of carotene from extracted oil of palm oil mill effluent (POME) was optimized by the simultaneous maximization of the carotene recovery in hexane fraction, carotene concentration and oil recovery. A central composite design consisting of 20 runs was then applied to study the factors of temperature, solvent amount and oil:adsorbent ratio. A second-order polynomial model was used to describe the experimental data regarding the responses. The main factors, solvent amount and oil:adsorbent ratio were the significant model terms for carotene and oil recoveries whereas oil:adsorbent ratio was the significant model term for carotene concentration. The optimum conditions for carotene recovery were predicted to be 40 degrees C, 200 mL solvent and 1:4.36 of oil to adsorbent ratio and the corresponding predicted values were 83.36% carotene recovery in hexane fraction, 23,628 ppm of carotene concentration and 100% oil recovery. The experimental values agreed with those predicted thus indicating the suitability of the model employed and the success of response surface methodology in optimizing the adsorption chromatography process. (C) 2010 Elsevier B.V. All rights reserved.